Augmentation schemes for particle MCMC
From MaRDI portal
Abstract: Particle MCMC involves using a particle filter within an MCMC algorithm. For inference of a model which involves an unobserved stochastic process, the standard implementation uses the particle filter to propose new values for the stochastic process, and MCMC moves to propose new values for the parameters. We show how particle MCMC can be generalised beyond this. Our key idea is to introduce new latent variables. We then use the MCMC moves to update the latent variables, and the particle filter to propose new values for the parameters and stochastic process given the latent variables. A generic way of defining these latent variables is to model them as pseudo-observations of the parameters or of the stochastic process. By choosing the amount of information these latent variables have about the parameters and the stochastic process we can often improve the mixing of the particle MCMC algorithm by trading off the Monte Carlo error of the particle filter and the mixing of the MCMC moves. We show that using pseudo-observations within particle MCMC can improve its efficiency in certain scenarios: dealing with initialisation problems of the particle filter; speeding up the mixing of particle Gibbs when there is strong dependence between the parameters and the stochastic process; and enabling further MCMC steps to be used within the particle filter.
Recommendations
- Augmented pseudo-marginal Metropolis-Hastings for partially observed diffusion processes
- scientific article; zbMATH DE number 7008320
- Particle MCMC algorithms and architectures for accelerating inference in state-space models
- Augmented Gaussian random field: theory and computation
- Particle Metropolis-adjusted Langevin algorithms
- On particle Gibbs samplers
- Particle rolling MCMC with double-block sampling
- The Calculation of Posterior Distributions by Data Augmentation
- On particle Gibbs sampling
Cites work
- A Bayesian analysis of some nonparametric problems
- Analytic Convergence Rates and Parameterization Issues for the Gibbs Sampler Applied to State Space Models
- Covariance structure of the Gibbs sampler with applications to the comparisons of estimators and augmentation schemes
- Efficient implementation of Markov chain Monte Carlo when using an unbiased likelihood estimator
- Ferguson distributions via Polya urn schemes
- Following a moving target -- Monte Carlo inference for dynamic Bayesian models
- scientific article; zbMATH DE number 2106098 (Why is no real title available?)
- MCMC for state-space models
- On disturbance state-space models and the particle marginal Metropolis-Hastings sampler
- On rates of convergence of stochastic relaxation for Gaussian and non- Gaussian distributions
- On some properties of Markov chain Monte Carlo simulation methods based on the particle filter
- On the efficiency of pseudo-marginal random walk Metropolis algorithms
- On-Line Inference for Hidden Markov Models via Particle Filters
- Optimal scaling for various Metropolis-Hastings algorithms.
- Particle Gibbs with ancestor sampling
- Particle learning and smoothing
- Particle learning for general mixtures
- Particle Markov chain Monte Carlo for efficient numerical simulation
- Particle Metropolis-adjusted Langevin algorithms
- Particle Metropolis-Hastings using gradient and Hessian information
- Rao-Blackwellization of Particle Markov Chain Monte Carlo Methods Using Forward Filtering Backward Sampling
- Remarks on the maximum correlation coefficient
- The Calculation of Posterior Distributions by Data Augmentation
- The pseudo-marginal approach for efficient Monte Carlo computations
- Uniform ergodicity of the iterated conditional SMC and geometric ergodicity of particle Gibbs samplers
Cited in
(11)- Conditional particle filters with diffuse initial distributions
- Improving performances of MCMC for nearest neighbor Gaussian process models with full data augmentation
- Full Bayesian inference in hidden Markov models of plant growth
- Augmented pseudo-marginal Metropolis-Hastings for partially observed diffusion processes
- A flexible particle Markov chain Monte Carlo method
- On particle methods for parameter estimation in state-space models
- scientific article; zbMATH DE number 7008320 (Why is no real title available?)
- Efficient data augmentation techniques for some classes of state space models
- A point mass proposal method for Bayesian state-space model fitting
- Conditional Particle Filters with Bridge Backward Sampling
- Particle MCMC and the correlated particle hybrid sampler for state space models
This page was built for publication: Augmentation schemes for particle MCMC
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q341152)
